Murray Stahl Q2 Letter: Shorting Path Dependent ETFs

Updated on

Murray Stahl Q2 letter to investors

FRMO Corp. Letter to Shareholders

Fiscal 2014




Yet, the modern notion of investing is the effort to control risk. Investors require portfolios to exhibit certain precise risk/reward tradeoffs. Securities in general, as they exist today, do not exhibit such characteristics. In order to achieve their risk preferences, investors have tried to blend various combinations of securities in portfolios, with the result—as a function of their persistent aggregate demand— that the various equity indexes of the world have been exhibiting increasing correlations with each other. This is also true for individual companies as well.


In the field of bonds, efforts have been made to disaggregate risks by so-called ‘tranching.’ The result was certainly less than satisfactory. One of the consequences was the creation of various tranches of sub-prime mortgage backed securities. The basic idea behind sub-prime and its various tranches was to appropriately price each tranche for risk by adjusting the yield.


However, there are two significant problems with this approach. The first is that yield is not merely compensation for the assumption of risk; it is also the price paid for capital by the person who wishes to access capital. If a given borrower is a so-called “high-risk” and might default on a 6% mortgage, surely the risk is magnified if the person in question must pay 9%. Hence, the question of risk pricing is reflexive; it is a paradox. If the bond is adequately priced, given the risk, it actually has the consequence of increasing the risk.


The second problem is that the various mortgage pools or tranches are composed of heterogeneous entities. Not all sub-prime borrowers actually do default. It is possible in theory to purchase a sub-prime tranche that has few, if any, defaults. It is also possible to purchase a sub-prime tranche with an enormous number of defaults. Yet, it is extraordinarily difficult to determine in advance what the default rate might be for any individual tranche. Consequently, since the default risk could not be quantified on a case-by-case basis, in practice the risk could not be controlled. In order to better achieve this result, it is necessary that the various elements be homogeneous, not heterogeneous. This quantification is not something easily done by an underwriter that wishes to earn a deal spread and that has an incentive to promote high volume issues.


Given modern computer technology, it is possible to better segregate risk categories. One day, exchanges will be places in which investors refashion portfolios to more precise risk/reward characteristics. The so-called buy side will predominate over the so-called sell side. The relationship between buyers of securities and exchanges might possibly be altered in currently unimaginable ways. Hence, we wish to retain an economic interest in one or more exchanges for entirely strategic reasons. This is not to say that our investment in the Bermuda Stock Exchange will not prosper for the customary reasons of increased volume and listing activity.

4)      Securities Sold Short. Our short positions consist primarily of path dependentindexes and ETFs. Although it might be rather hard to believe, the formal definition of a path dependent index is an index or ETF that will ultimately achieve a price of virtually zero. We wish to pursue this investment since we can use our existing assets as collateral, and no investment as such is necessary since it is a short sale.


As of fiscal year end, our short sale position had a cost basis of $5,634,323 and a market value of $1,709,985. Hence, we now have an unrealized profit of $3,924,338, earned without the actual deployment of firm capital (aside from nominal financing costs), although it must certainly be said that our capital was and remains at risk.


Path dependent indexes and ETFs experience net asset value decay due to several factors. First among these is if the ETFs in question are commodity oriented. The question of contango versus backwardation arises. Generally, commodities are in contango position. This means that the forward months trade at higher prices than the current or spot months. Let us assume that a given commodity ETF represents a one-month future. Of course, the ETF can purchase a one-month future. The problem is that after one day the ETF no longer represents a one-month future, since that contract is one day closer to expiration. It will be a 29-day future. However, since it is an index it must be formulaically consistent. Ergo, it will sell one-thirtieth of its one-month-less-one-day future and use the money to buy a one-thirtieth position in the two-month-less-one-day future. It will then still have, on a weighted basis, a one-month future position. This forward contract is more expensive if the commodity in question is in contango. In this manner the ETF constantly repeats the process and earns that which is called in the futures industry negative roll yield. If one wishes to earn this yield, one establishes a short position in the product that engages in this practice.


Another source of decay is leverage. There are many 2X and 3X leveraged ETFs. The leverage itself ultimately becomes a source of decay, since the formulaic constraint placed upon the ETF results from its identity in an index that must have the same leverage ratio on each trading day.


An example, albeit extreme, will make this clear. Let us assume that a given investor wishes to purchase the S&P 500 on leverage. In this case the investor will use 2X leverage. A $10,000 investment in equity will be added to $10,000 of borrowings for a $20,000 position in the S&P 500. The investor’s position on day one is as follows:

$10,000 initial deposit equity + 10,000 margin loan

=  $20,000 gross exposure


Let us now assume for illustrative purposes only that the S&P 500 doubles in value. The position, if this occurs, is as follows:


$10,000 initial deposit equity

+      10,000 margin loan

+       20,000 unrealized gain = $40,000 gross exposure


–      10,000 margin loan

= $30,000 new equity


If the S&P 500 were then to lose 50% in value, the $40,000 in gross exposure would become $20,000 and the investor would unfortunately, in this example, return to the original position. This would be as follows:


$10,000 initial deposit equity + 10,000 margin loan

= $20,000 gross exposure


In contradistinction, here is what occurs when the same strategy is attempted in an index ETF format. It will be recalled that the example is deliberately extreme for illustrative purposes. In this case an investor places $10,000 in an ETF, perhaps based on the S&P 500, and leveraged 2X. Initially, the position is no different from that of an individual investor, with the exception that the leverage is applied within a fund context.


$10,000 equity + 10,000 loan

=  $20,000 gross exposure


Let us now assume that in one day the S&P 500 were to double, as preposterous as this assumption will seem. The position at the end of the trading day is identical to that of the individual investor.


$10,000 equity

+        10,000 loan

+         20,000 unrealized appreciation = $40,000 gross exposure


– 10,000 margin loan = $30,000 net equity

However, since the fund in question is an index, it must have the same exposure for each trading day. Consequently, since the fund now has $30,000 of net equity, it must leverage 2X and will purchase, with new debt, another $20,000 of the S&P 500. Its position now become the following:


$30,000 equity + 30,000 debt


=  $60,000 gross exposure


Let us now assume that the S&P 500 were to decline by 50% in one day. A 50% loss on $60,000 of gross exposure would result in $30,000 of gross exposure. Hence there would be no remaining equity. The position at this point is therefore as follows:


$ 0 equity + 30,000 debt

= $30,000 gross exposure


The investor has lost 100% of the equity merely as a consequence of the underlying index rising, then falling, by the same 50%. The example is deliberately extreme to avoid using calculus. However, in slower motion this is essentially the path dependent phenomenon. The constancy of leverage necessary to maintain legal status as an index essentially guarantees that the maximum leverage will be used at the high point of the index. Maximum should be understood in this context to mean that the leverage is maximized in dollar, not percentage terms. Of course, the maintenance of maximum leverage at a market high makes possible maximum portfolio damage. Those readers who recall the concept of maxima from calculus can perhaps visualize this circumstance in a Cartesian plane.


In reality, the market variability is not nearly this extreme. Nevertheless, the leverage doubles the market variability and the portfolio must constantly trade to maintain an invariant leverage position. This trading exacts its inevitable toll on the portfolio in terms of transaction costs.


One final minor point is interesting in this regard. When one is short the ETF, the management fee charged against the ETF helps to depreciate the investment, which is the purpose of a short position. Hence, short positions on ETFs actually enable the short seller to effectively “earn” the management fee.


It is for this reason that we maintain short positions in path dependent ETFs. In the short run, the strategy is not devoid of mark-to-market risk. However, in the intermediate run and longer, this should prove to be a highly productive investment.

Full letter in PDF here via Murray Stahl FRMO Corp. Letter to Shareholders

Via FRMO Corp

Leave a Comment